Revista Ingeniería de Construcción RIC

Vol 32 Nº3 2017 www.ricuc.cl

ENGLISH VERSION.....................................................................................................................................................................................................................................................

Análisis sísmico de gradas con tres modelos mediante los

métodos: espectral y del espectro de capacidad, que sirven de
acceso a una estructura con aisladores sísmicos
Seismic analysis of stairs with three models using the spectral
and the capacity spectrum methods, give access to a building
with seismic isolators
1
R. Aguiar *, D. Mora **, D. Muñoz ***, S. Serrano ***, D. Vera ***, M. Rodríguez *, D. Páez ***

* Universidad de Fuerzas Armadas ESPE, Quito. ECUADOR

** Escuela Politécnica Nacional, Quito. ECUADOR
*** Universidad Laica Eloy Alfaro de Manabí, Manta. ECUADOR

Fecha de Recepción: 26/01/2017

Fecha de Aceptación: 22/09/2017
PAG 113-130

Abstract

At the Universidad de las Fuerzas Armadas - ESPE, in Ecuador, different structural blocks with and without seismic isolators of triple friction pendulum are being built.
In one of them, the access stairs to the floors is located in a Structural Block with no seismic isolators; therefore, it has special construction details to allow a free
displacement of the Isolators’ Building without impacting the Stairs’ Building at all. This paper addresses these details. On the other hand, three models (each with
different degrees of freedom) for seismic analyses of a longitudinal frame and a transversal frame are shown. The results were obtained and compared using the
Spectral Method and the Capacity Spectrum Method. Moreover, these results highlight the importance of appropriately selecting the numerical calculation model.
Later on, the seismic analysis of the Structural Block of the Access Stairs is shown, using a three-phase constitutive model.

Keywords: Spectral Method; Capacity Spectrum Method; Seismic Isolators; Access Stairs.

Resumen

En la Universidad de las Fuerzas Armadas ESPE de Ecuador, se están construyendo algunos Bloques Estructurales con aisladores sísmicos de triple péndulo de
fricción. En uno de ellos, las gradas de acceso a los diferentes pisos se encuentran en un Bloque Estructural sin aisladores sísmicos, por lo que tiene detalles
especiales en su construcción para permitir el libre desplazamiento de los aisladores sin posibilidad de impacto con el Bloque de Gradas. En este artículo se indican
esos detalles. Por otra parte, se presentan tres modelos con diferentes grados de libertad para el análisis sísmico de un pórtico en sentido longitudinal y de un pórtico
en sentido transversal de la grada; se compara sus respuestas empleando el Método Espectral y el Método del Espectro de Capacidad. Destacando con ello la
importancia de seleccionar en forma apropiada el modelo numérico de cálculo. Posteriormente se presenta el análisis sísmico del Bloque Estructural al que dan
acceso las gradas, utilizando como constitutiva un modelo de tres fases.

Palabras clave: Método espectral, método del espectro de capacidad, aisladores sísmicos, gradas de acceso

1. Introduction
The Universidad de las Fuerzas Armadas – ESPE,
located in the Valley of Los Chillos, is building eight structural
blocks that will house the Research and Postgraduate Centers.
Structural Blocks 1, 2, 3, 6, 7 and 8 have seismic isolators of
triple friction pendulum (TFP) and the 4 and 5 blocks have
The emergency stairs was built with steel square
tubular columns of 250/250/20 mm and “I” beams of
400/200/15/12 mm, in Structural Blocks without seismic
isolators, which also contain the elevator structure. The circles
in Figure 1 show the emergency stairs analyzed in the present
th th
paper, which gives access to the 1 and 7 Structural Block. st th

Shear Link energy dissipators. Figure 1 shows an overview of Figure 2a) shows the structural block without isolators,
the mentioned Structural Blocks, which total a construction containing the stairs and the elevator. The seismic analysis of
area of approximately 23,500 square meters. the J-Frame (longitudinal) and the 7’ Frame (transverse) will
be described further on.
On the other hand, Figure 2b) shows the development
of the stairs, where to go up from one floor to another there
are three inclined segments and two landings between axes J
– J”. The elevator is located between axes J’ – J”’ and axes 7a
– 7b. There is a room on the back of the elevator (axes J – J’),
1
Corresponding author: on the right side (axes 7b – 6a).
Head Professor of the Department of Earth Sciences and Construction The entrance to one of the Structural Blocks with
E-mail: rraguiar@espe.edu.ec isolators, for example to Block 7, is from axis I. The problem
to be solved is how to build the joint between the 7 th

Structural Block with TFP isolators, which will show a

Revista Ingeniería de Construcción Vol 32 Nº3 Diciembre de 2017 www.ricuc.cl 113

considerable lateral displacement (60 cm GAP), and the
Stairs Block indicated in Figure 2.
The solution given to prevent the impact of these two
structures during an earthquake is shown in Figure 3a) for the
stairs and in Figure 3b) for the elevator. The stairs shows that
the last step is the cantilever beam of the Structural Block 7.
The steel steps reach 7 cm, except those of the slab of Block
7, so that they can move independently without impact. In
the upper part of a floor, the first step is also located at 7 cm
of the slab; see Figure 4. It is important to note that the step
moves independently from the Isolated Block, so that they do
not impact with each other; therefore, the deck has two slabs
at different heights.
Figure 3b) shows the access from the elevator to the
Block with isolators. Here, there is a cantilever beam at a
lower level than the isolated structure; on top of the beam
there is a flexible foam filling, so that people coming out of
the elevator can walk at a single level. During an earthquake,
the filling is expected to break, so that the isolated Block can
have a complete displacement. Evidently, the flexible foam
will be replaced after the earthquake, if necessary.

Figure 1. Overview of the Structural Blocks, where the UFA-ESPE Research and Postgraduate Centers will operate

a) b)

Figure 2. Plan view of the Structural Block without isolators: a) Implementation of Columns; b) Stairs Development

Revista Ingeniería de Construcción Vol 32 Nº3 Diciembre de 2017 www.ricuc.cl

114
 Revista Ingeniería de Construcción RIC
Vol 32 Nº3 2017 www.ricuc.cl
ENGLISH VERSION.....................................................................................................................................................................................................................................................

a) b)

Figure 3. Detail of the entrance to Structural Block 7 with seismic isolators: a) Access by stairs; b) Access by elevator.
Source: Aguiar and Pazmiño (2016)

Figure 4. Construction detail of the steps to or from the Structural Block with isolators

2. Numerical models for analyzing the

longitudinal frame

Apparently, stairs are a very simple structure, so that Figure 5b) shows the degrees of freedom for the
sometimes the seismic analysis is even ignored or it is made model associated to Figure 5a), which has 10 lateral
with extremely elemental models, which do not report coordinates and 55 degrees of freedom in total. Therefore, a
adequate results. Therefore, a seismic analysis of the stairs direct assembly of 55 by 55 will calculate the stiffness matrix,
from the Access Block to the 7 Structural Block was defined. th
which is condensed to a matrix of 10 by 10 for a seismic
A 2D seismic analysis is applied to the longitudinal J- analysis (Model M10).
Frame and the transverse 7’ frame. This section describes the Regarding the simplified model analysis of Figure 6a),
J-Frame; therefore, Figure 5a) indicates the frame geometry Figure 6b) shows 7 lateral coordinates and 40 degrees of
and Figures 6a) and 7a) present two calculation approaches, freedom; thus, the lateral stiffness matrix associated to the
which are actually used to simplify the analysis, so as not to lateral coordinates is 7 by 7 (Model M7).
work with inclined elements, but as we will see, it conduces Finally, the third model presented in Figure 7a) has 4
to different results. lateral coordinates and 31 degrees of freedom; see Figure
7b), (Model M4).

Revista Ingeniería de Construcción Vol 32 Nº3 Diciembre de 2017 www.ricuc.cl 115

 a) b)

Figure 5. a) Geometry and loads of J-Frame, b) Model with 10 primary coordinates

a) b)

Figure 6. a) Geometry and loads of J-Frame, b) Model with 7 primary coordinates

Revista Ingeniería de Construcción Vol 32 Nº3 Diciembre de 2017 www.ricuc.cl

116
 Revista Ingeniería de Construcción RIC
Vol 32 Nº3 2017 www.ricuc.cl
ENGLISH VERSION.....................................................................................................................................................................................................................................................

a) b)

Figura 7. a) Geometry and loads of J-Frame, b) Model with 4 primary coordinates

Annex A includes the theoretical frame of the Spectral
Method. In both cases, the seismic analysis is the one
recommended by the Ecuadorian Construction Standard
NEC-15 for a seismic region whose PGA is 0.4g for a type-C
ground.
3. Results with the Spectral Method
Table 1 shows the analysis model; the fundamental
period of vibration T, the maximum lateral displacement q,
the floor drift y, and the base shear V in the J-Frame, using the
Spectral Method, with a seismic force reduction factor R = 4.
It can be seen that the model with 10 degrees of freedom
M10 reports the smallest period, and the same occurs with
the lateral displacement. The lowest floor drift occurs with
model M7 and the lowest shear, with model M4.
Figure 8a) shows the lateral displacements with each
model, and Figure 8b) shows the floor drifts. The Model M4 is
the most simple, but it should not be used because it reports
very high values and it also differs from the geometry of the
stairs. It is feasible to work with Model M7, but it is best to
use M10.
0,

Table 1. Results of the seismic analysis of the J-Frame. Spectral Method

Model 𝑇 𝑞 𝛾 𝑉!
(s.) (cm) (%) (T.)
M10 0.483 9.57 1.5 5.36
M7 0.542 12.32 1.3 5.64
M4 0.806 16.13 1.9 4.46

Revista Ingeniería de Construcción Vol 32 Nº3 Diciembre de 2017 www.ricuc.cl 117

 a)
Figure 8. a) Lateral Displacements b) Floor Drifts, of J-Frame
4. Results with the Capacity Spectrum
Method
The graphs at the left of Figure 9a show the seismic-
resistant capacity curve that relates the base shear and the
displacement on the top, through a monotonic pushover.
Then, the performance point was found; therefore, the center
part of Figure 9 shows the structure’s capacity spectrum
(green line), the design spectrum of NEC-15 (red line) and
the point that allows finding the design point (blue line) when
applying the Capacity Spectrum Method described in Annex
B.
Finally, the graphs at the right of Figure 9 correspond
to the damage level that the structure will suffer during the
earthquake. The green color means that the section exceeded
the plastic moment M and entered the nonlinear range. Other
Y
structures may also present a red color, which means that a
plastic hinge was formed in the section, but it was not the
case in the studied structure.
Figure 9a) corresponds to the analysis with the model
of 10 degrees of freedom M10, where the structure works
elastically, because no colors are observed on the graph to
the right.
Revista Ingeniería de Construcción
118
b)
Figure 9b) corresponds to the model M7, analyzed
with 7 degrees of freedom, where the structure does not
suffer any damage either. Lastly, Figure 9c) shows the
performance with the model M4. In this case, there are two
sections that have exceeded the plastic moment, which
correspond to the central column in the upper part and the
beam located on the right of the first floor. This is because the
model M4 reports big displacements.
Table 2 presents the results obtained with the Capacity
Spectrum Method in the J-Frame, and indicates the analysis
model, the displacement corresponding to the design point
(q) and the associated acceleration (A), the displacement on
the top of the structure (q ), the floor drift y and the base
T
shear (V ). The resulting displacements and floor drifts are
0
similar to those found with the Spectral Method, indicated in
Table 1. The base shear is different, because the shear in
Table 1 has been divided for R = 4, in the Spectral Method,
and now there is an elastic base shear.
Vol 32 Nº3 Diciembre de 2017 www.ricuc.cl
 Revista Ingeniería de Construcción RIC
Vol 32 Nº3 2017 www.ricuc.cl
ENGLISH VERSION.....................................................................................................................................................................................................................................................

Table 2. Results of the seismic analysis of J-Frame. Capacity Spectrum Method

Model 𝑞  (𝑚) 𝑚 𝑞!  (𝑚) 𝛾  (%) 𝑉!  (𝑇)

𝐴   ! ! !
1gdl 𝑠
M10 0.075 11.033 0.105 1.739 23.93
M7 0.103 11.565 0.134 1.762 24.06
M4 0.153 9.259 0.179 2.271 20.12

a)

b)

c)

Figure 9. Seismic-resistant capacity, Capacity Spectrum Method, and zones with damage on J-Frame, a) Model M10, b)
Model M7, c) Model M4

Revista Ingeniería de Construcción Vol 32 Nº3 Diciembre de 2017 www.ricuc.cl 119

5. Numerical models for analyzing the
transverse frame
Figures 10 to 12 present the geometry, loads, sections and 4 degrees of freedom, respectively, in the presence of a
of the elements for the 7’ Frame, and the analysis models design basis earthquake (DBE).
M10, M7 and M4, with a seismic analysis considering 10, 7

a) b)

Figure 10. 7’ Frame a) Geometry and loads, b) Model with 10 primary coordinates

a) b)

Figure 11. 7’ Frame a) Geometry and loads, b) Model with 7 primary coordinates

Revista Ingeniería de Construcción Vol 32 Nº3 Diciembre de 2017 www.ricuc.cl

120
 Revista Ingeniería de Construcción RIC
Vol 32 Nº3 2017 www.ricuc.cl
ENGLISH VERSION.....................................................................................................................................................................................................................................................

a) b)

Figura 12. 7’ Frame a) Geometry and loads, b) Model with 4 primary coordinates

Table 3 indicates the results of the seismic analysis of Figure 13 shows the displacements and floor drifts
the 7’ Frame. Once again, the results of model M4 are obtained with models M10, M7 and M4 for the 7’ Frame
different compared with the other models and the best is to
work with model M10.

Table 3. Results of the seismic analysis of the 7’ Frame

Model 𝑇 𝑞 𝛾 𝑉!
(s) (cm) (%) (T)
M10 0.597 14.00 1.30 3.66
M7 0.579 13.46 1.48 3.96
M4 0.652 14.51 1.48 3.53

a) b)

Figure 13. a) Lateral displacements and b) Floor drifts of the 7’ Frame

Revista Ingeniería de Construcción Vol 32 Nº3 Diciembre de 2017 www.ricuc.cl 121

6. Results with the capacity spectrum
method

The results obtained with the Capacity Spectrum should be noted that in model M10, the final knot from a
Method in the 7’ Frame, with models M10, M7 and M4, are column of the first floor enters the nonlinear range (green
indicated in Figure 14 and their maximum values are color).
indicated in Table 4. Likewise, the model M4 is not reliable. It

a)

b)

c)

Figure 14. Seismic-resistant capacity, Capacity Spectrum Method and damaged zones on the 7’ Frame, a) Model M10,
b) Model M7, c) Model M4

Revista Ingeniería de Construcción Vol 32 Nº3 Diciembre de 2017 www.ricuc.cl

122
 Revista Ingeniería de Construcción RIC
Vol 32 Nº3 2017 www.ricuc.cl
ENGLISH VERSION.....................................................................................................................................................................................................................................................

Table 4. Results of the seismic analysis of the 7’ Frame. Capacity Spectrum Method

Model 𝑞  (𝑚) 𝑚 𝑞!  (𝑚) 𝛾  (%) 𝑉!  (𝑇)

𝐴   ! ! !
1gdl 𝑠
M10 0.114 11.238 0.153 1.507 16.41
M7 0.092 11.18 0.129 1.417 15.34
M4 0.127 11.089 0.162 1.709 15.99

Again, the results reported by the Capacity Spectrum be built with a construction joint, since the 7 Block has th

Method are quite similar to those reported by the Spectral anirregular elevation. However, if they have seismic isolators,
Method at the level of lateral displacements and floor drifts. In in this case the triple friction pendulum (TFP), the best is to
order to compare the shears, the values reported by the build without joints, because, on the one hand, the column
Spectral Method should be multiplied by R = 4. axis is eliminated with its respective isolators that should be
put in the joint. On the other hand, it avoids using special
pipes to pass plumbing or electrical facilities from one block
to another. Finally, if a construction joint is used, there is a
7. Seismic analysis of structural blocks problem to be solved regarding the approximately one-meter
with Isolators gap on the façade, which is caused when isolators are
installed and two independent blocks are built.
Therefore, the best is a single Structural Block, referred
The analyzed Stairs Block gives access to the 6 and 7 th th

to as 6-7, where TFP isolators are placed on top of reinforced-

Structural Block on the cantilever side of 3.51 m, located on concrete columns of 80/80 cm, on the basement side and
the W-axis in Figure 15. The entrance is located between axes above the pedestals of 120/120 cm, coming out from the
39 and 41. Bays 32-33 and 33-36 have a basement; plinths in axes 39, 41 and 43. The isolators have a length of
therefore, initially, the idea was to build this sector approximately one meter; therefore, a basket of 120/120 cm
independently from the other bays. These two bays formed was built on top of the 80/80 cm columns, as shown on the
the 6th Building and the rest formed the 7th Building left side of Figure 16, and the isolators for the three
If the structure has no isolators, Blocks 6 and 7 should mentioned axes are shown on the right side.

Figure 15. Structural Block 6 and 7 with TFP isolators, where the stairs are connected to bays 39-41 on the cantilever side of 3.51
m. Source: Aguiar and Pazmiño (2016)

Revista Ingeniería de Construcción Vol 32 Nº3 Diciembre de 2017 www.ricuc.cl 123

 Figure 16. TFP isolator on top of concrete columns and on top of the pedestal in Block 6-7. Source: Aguiar and Pazmiño (2016)
Annex C addresses the theoretical frame of the three-
phase model proposed by McVitty and Constantinou (2015),
concerning the force-displacement constitutive curves of
isolators, which was used in the seismic analysis with the
Spectral Method. Figure 17 indicates the displacements in the
X-direction and the Y-direction for design basis earthquakes
DBE (obtained with NEC-15, for a return period of 475 years)
and maximum considered earthquakes MCE (obtained by
multiplying the DBE by 1.4 and it has a recurrence interval of
1,000 years, Constantinou et al., 2016).
Displacements in the Y-direction are towards axes 8’,
7’, 7a and 6a of the access stairs. It is observed that, in the
Revista Ingeniería de Construcción
124
event of a MCE, maximum lateral displacements of
approximately 47 cm are expected in the isolated block.
Figure 17 also shows lateral displacements considering
the non-isolated structure, which seems to have almost the
same displacement or a bit less, but there is a great
difference: in the structure with isolators, the isolator shows
the greatest displacement (horizontal line) and the structure
above shows a very small displacement in relation to the
isolators’ displacement.
Thus, the structure without isolators is damaged,
because it suffers important displacements, while the
structure with isolators has relatively small displacements.
Vol 32 Nº3 Diciembre de 2017 www.ricuc.cl
 Revista Ingeniería de Construcción RIC
Vol 32 Nº3 2017 www.ricuc.cl
ENGLISH VERSION.....................................................................................................................................................................................................................................................

Figure 17. Lateral displacements in the Structural Block with isolators, in the presence of DBE and MCE in two directions.
Source: Aguiar et al., 2016.

8. Remarks and Conclusions

seismic isolators This paper has presented a The seismic analysis was based on the Spectral
seismic analysis of a 4-floor stairs Method and the Capacity Spectrum Method, and both
designed without , which gives access to a structural block methods presented quite similar results. In the professional
with triple friction pendulum seismic isolators. The analysis practice, the Spectral Method is preferred due to its simplicity.
considered three models with 10 degrees of freedom (M10), Instead, the Capacity Spectrum Method is more elaborated,
7 degrees of freedom (M7) and 4 degrees of freedom (M4) but it shows how structural members will look after the
respectively. It was concluded that the latter is the simplest earthquake, if they suffered any damage or not.
model, but it does not report reliable results.

Revista Ingeniería de Construcción Vol 32 Nº3 Diciembre de 2017 www.ricuc.cl 125

 Finally, the study presents the results obtained from the system of linear equations to be solved. The matrix
the seismic analysis of the Structural Block with isolators, to equation (5) indicates the resulting stiffness matrix following
which stairs give access to, which considered the isolators’ the triangularization to find the lateral stiffness matrix.
constitutive model of McVitty and Constantinou (2015). A
spatial analysis with the Spectral Method was carried out in 3. In the 2D case, the mass matrix 𝑴 is diagonal and it
floor coordinates, in the presence of DBE and MCE. Results is composed by the total masses of each floor.
show that maximum displacements of approximately 47 cm
are expected, thereby highlighting that it is the isolators that 4. The problem of own values and vectors is solved.
will have the greatest displacement, around 40 cm. Therefore,
the relative displacements of the structure in relation to the
isolators are around 7 cm, which means that they will not be 𝑲𝑳 −  𝝀  𝑴  𝝓 = 𝟎 (6)
damaged.
Since the isolated structure has big displacements, it Where 𝜆 are the own values for finding the vibration
was made sure that the Structural Block of the Stairs, which is frequencies 𝑊! and the period 𝑇, in each vibration mode 𝜙.
made of steel, was a rigid structure; however, this causes
!  !
displacements of around 15 cm. Consequently, constructive 𝑊!" =   𝜆!                                                          𝑇! = (7)
!!"
preventive measures were taken in the access of the Stairs
Block and the Elevator to the Block with seismic isolators,
5. The spectral acceleration 𝐴!" is found with each
which have been presented herein.
period in the inelastic spectrum. A reduction factor
All these experiences regarding structural and
R = 4 was used in the stairs structure.
construction analyses are expected to be beneficial for those
who are working in the design of buildings with seismic
6. The modal participation factors Γ are determined in
isolators.
each mode.

9. Annex A: Analysis procedure with 𝛤! =

𝝓 𝒊 𝒕  𝑴  𝒃
(8)
𝝓 𝒊 𝒕  𝑴  𝝓(𝒊)
the spectral method
Where 𝒃 is the incidence vector of the ground motion in the
degrees of freedom. For the 2D case, the vector 𝒃 is a unit
1. The degrees of freedom are properly numbered,
vector.
thereby grouping primary coordinates and
secondary coordinates. Then, the stiffness matrix of
7. Modal displacements 𝒒(𝒊) are found
the structure is obtained by direct assembly (Lamar
and Fortoul, 2007).
!! !
𝒒(!) =  𝐴!"  Γ!    𝝓(𝒊) (9)
!!
𝑲 𝑲𝒂𝒃
𝑲 = 𝒂𝒂 (1)
𝑲𝒃𝒂 𝑲𝒃𝒃 Now, a modal combination criterion should be used to find
the resulting displacements. This paper applied the modal
combination method of Complete Quadratic Combination
Where 𝑲 is the stiffness matrix of the structure, which is (CQC). Chopra (2001).
partioned according to the number of primary coordinates in
the submatrices 𝑲𝒂𝒂 𝑛𝑎, 𝑛𝑎 ;   8. The inelastic displacements 𝒒!"# are determined by
𝑲!" 𝑛𝑎, 𝑛𝑏 ;  𝑲!" = 𝑲!!" ;  𝑲!! 𝑛𝑏, 𝑛𝑏 . Where 𝑛𝑎, 𝑛𝑏, is the multiplying the elastic displacements 𝒒 by the
number of primary and secondary coordinates. response reduction factor 𝑅.

2. The stiffness matrix associated to the lateral 9. Finding the inelastic floor drifts 𝛾 (!) .
coordinates is determined, and there are three
options available: the first is to find the inverse of the
matrix (Equation 2), the second is to solve linear 𝛾 (!) =  
!!"#  !   !  !!"#  !!!
(10)
equations (Equations 3 and 4) and the third is !!

through the Gauss triangularization method

(equation 5). For the latter, primary coordinates are Where 𝑞!"#  ! , 𝑞!"#  !!! , are the inelastic displacements in the
numbered at the end. floors 𝑖, and 𝑖 − 1; ℎ! is the floor height.

𝑲! = 𝑲!! − 𝑲!"   𝑲!! !𝟏  

𝑲!" (2) 10. The forces in each mode 𝐹 (!) are determined, and
the modal combination criterion CQC is applied for
𝑲𝑳 =   𝑲!! + 𝑲!"  𝑻   (3) the resultant.

𝑲!!  𝑻 =   −  𝑲!" (4)

𝐹 (!) = Γ!  𝐴!"  𝑴  𝝓(𝒊)   (11)
𝟎 𝑰 −𝑻 𝒒!
=   (5) All calculations indicated above have been made in the
𝑸𝒃 𝟎 𝑲 𝑳 𝒒!
present paper. Controls based on the seismic standard of
Where 𝒒! , 𝒒! , are the vectors containing the displacements minimum base shear, the 𝑃 − 𝛥 effect, and calculating the
and rotations (generalized coordinates). Equation 4 defines

Revista Ingeniería de Construcción Vol 32 Nº3 Diciembre de 2017 www.ricuc.cl

126
 Revista Ingeniería de Construcción RIC
Vol 32 Nº3 2017 www.ricuc.cl
ENGLISH VERSION.....................................................................................................................................................................................................................................................

accidental torsion, are still pending. Barbat and Bozzo

(1997), Chopra (2014). !
𝑬𝑰 = (13)
∅

Where E is the Young’s Modulus, I is the moment of Inertia of

10. Annex B: Calculation procedure the analyzed section, M is the yield Moment and ∅ is the
with the capacity spectrum Curvature.
method 4. After several cycles of small lateral loads,
accumulating resulting forces, eventually several
The Capacity Spectrum Method requires three primary sections will lose the capacity to resist moments,
elements: capacity, demand (displacement) and performance which is known as a plastic hinge. As mentioned
(ATC-40, 1996). earlier, the structural analysis is based on stiffness
and transformation matrices. When a plastic hinge is
Capacity formed, the problem is that the stiffness matrix of the
The capacity curve of a structure is represented by a structure becomes singular and, when calculating
pushover curve (nonlinear static analysis). The most the displacements and rotations of each degree of
convenient way to represent it on a graph is by the base shear freedom, it is necessary to know the inverse of the
and the displacement at the top of the building. stiffness matrix, which is impossible.
A computer process is needed to achieve severe
iterations of static analyses and define the capacity curve (Ger 𝒒 = 𝑲\𝑸 (14)
and Cheng, 2012).

Process: Where K is the Stiffness matrix of the structure, Q is the load

Vector in degrees of freedom, q is the displacement or
1. Definition of the geometry of the structure, the rotations in the degrees of freedom.
sections of each element and the degrees of freedom This problem can be solved by creating new stiffness
of each knot. matrices for elements and transformation for different systems
2. Calculation of the actions resulting from gravity of plastic hinge formations.
loads.
3. Then, they are accumulated to the actions resulting 5. Iterations will stop only when the structure
from lateral loads. collapses. The collapse criterion may include floor
drifts greater or equal to 2% (ASCE 7, 2010; NEC,
The structure is loaded with relatively small lateral 2015) or a global drift enough to obtain results at
forces. A static analysis is made in each iteration, the performance point; generally, at a 4% limit, the
accumulating the resulting actions for the next global drift is based on the total height of the
iteration. Sections will eventually flow, reducing building.
their stiffness 𝐸𝐼 (Young’s Modulus by the moment Moreover, the software used herein will stop if the stiffness
of Inertia of the section). matrix of the structure is singular, because this indicates
that the structure is unstable; this situation occurs when
It is possible to know the nonlinear behavior of the several plastic hinges are formed.
structure through the variation of the 𝐸𝐼 stiffness,
because this parameter is involved in the static Demand
analysis process. It is the ground motion during an earthquake. For the
nonlinear method, it is easier and more direct to use a set of
4𝐸𝐼/𝐿 2𝐸𝐼/𝐿 0 lateral displacements as a design condition. For a given
𝒌𝟏 = 2𝐸𝐼/𝐿 4𝐸𝐼/𝐿 0 (12) structure and ground motion, the displacement demand is an
0 0 𝐸𝐴/𝐿 estimate of the maximum considered response of a structure
during the ground motion. The codes allow obtaining the
Where k1 is the stiffness matrix of a section in reduced design spectrum, which relates the response acceleration of
local coordinates; 𝐸𝐼 is the flexural stiffness that changes the structure with the fundamental period of the structure.
according to the damage degree of the section; L is the length
Performance
of the element; EA is the axial stiffness.
In order to verify the performance of the structure, the
Here, actions were calculated by the structural
structure capacity curve and the seismic demand spectrum
analysis method of stiffness, using stiffness and transformation
curve must be placed in the same format: displacement with
matrices for reduced, local and generalized systems of each
spectral acceleration (Chopra and Goel 2000, 2001).
section, and assemblies based on the degree of freedom of
The performance must be checked once the capacity
each element, in order to obtain the stiffness matrix of the
and demand curves have been defined (ASCE 41, 2013). A
whole structure. Subsequently, the deformations of each
performance check verifies that the structural and non-
degree of freedom and the resulting actions are found.
structural components are not damaged beyond acceptable
In order to know the flexural stiffness 𝐸𝐼 in every stage
limits.
of a section, it is necessary to know the moment–curvature
diagram of every section, since the hypothesis of Bernoulli
and Navier shows that:

Revista Ingeniería de Construcción Vol 32 Nº3 Diciembre de 2017 www.ricuc.cl 127

 The performance points represent the maximum
considered displacement in the design earthquake, given the
fact that the mathematical model directly considers the
inelastic effects the material’s response. The calculated
internal forces are reasonable approximations expected during
the design earthquake.
Figure A1, which shows the geometry and nomenclature of
the TFP isolator, considering the dimensions used in the
constructions of the Universidad de las Fuerzas Armadas -
ESPE, in Ecuador.
The following equations are necessary to describe the
three-phase model:

11. Three-phase constitutive model 𝑅!,!"" = 𝑅! − ℎ! Para i de 1 a 4. (15)

!!,!""
𝑑!∗ =   (16)
There are different constitutive models dealing with !!
the nonlinear behavior of triple friction pendulum isolators
(TFP), among which the following can be highlighted: the Where R is the radius of curvature; h is the height, indicated
i i

equivalent linear method of ASCE 7-10; the 5-phase model in Figure A1; R is the effective radius of curvature; 𝑑!∗ is the
i,e f f

proposed by Fenz and Constantinou (2007, 2008), and the displacement capacity. The subscript i varies from 1 to 4. The
three-phase model used in this study, proposed by McVitty 3 phases or regimes of the model are the following:
and Constantinou (2015). The latter is indicated below in

Figure A1. Geometry of the triple friction pendulum isolator FTP8833 / 12-12 / 8-6, used in the construction of the
Universidad de las Fuerzas Armadas - ESPE.

Regime 1

Here, the displacement occurs only in plates 2 and 3. Where u is the lateral displacement of the isolator; F is the
applied lateral force; W is the weight on the isolator. The left
0   ≤ 𝑢   ≤   𝑢 ∗ side of Figure A2 shows the inner isolator sliding along
surfaces 2 and 3; the right side shows the corresponding
𝑢 ∗ = 2   𝜇! − 𝜇!  𝑅!,!"" (17) hysteresis diagram.

!
𝐹=  𝑢 +   𝜇!  𝑊 (18)
!  !!,!""

Figure A2. Isolator Performance in Phase I. Source: McVitty and Constantinou (2015)

Revista Ingeniería de Construcción Vol 32 Nº3 Diciembre de 2017 www.ricuc.cl

128
 Revista Ingeniería de Construcción RIC
Vol 32 Nº3 2017 www.ricuc.cl
ENGLISH VERSION.....................................................................................................................................................................................................................................................

Regime 2
The movement in the inner isolator reaches the stops 𝑢 ∗   ≤ 𝑢   ≤  𝑢 ∗∗
and starts sliding along surfaces 1 and 4. Normally, in the
event of an earthquake of medium to high intensity, the 𝑢 ∗∗ = 𝑢 ∗ + 2  𝑑!∗ (19)
isolator works in this regime. The following equations and
!
Figure A3 show the corresponding hysteresis curve. 𝐹=   𝑢 − 𝑢 ∗ +   𝜇!  𝑊 (20)
!  !!,!""

Figure A3. Isolator Performance in Regime II. Source: McVitty and Constantinou (2015)

Regime 3
This regime occurs when the earthquake is extremely 𝑢 ∗∗   ≤ 𝑢   ≤   𝑢!"#
strong and the inner isolator hits the outer stops; in these
conditions, the inner isolator begins to slide along surfaces 2 𝑢!"# = 2  𝑑!∗ + 2  𝑑!∗ (21)
and 3. Figure A4 shows the calculation equations and the
performance. ! !
𝐹=   𝑢 − 𝑢 ∗∗ +     𝑢 ∗∗ −   𝑢 ∗ +   𝜇!  𝑊 (22)
!  !!,!!! !  !!!""

Figura A4. Isolator Performance in Regime III. Source: McVitty and Constantinou (2015)

12. References
Aguiar R., Pazmiño E. (2016), Detalles constructivos debido a los aisladores sísmicos en las construcciones de la Universidad de Fuerzas
Armadas ESPE en Ecuador, Revista Internacional de Ingeniería de Estructuras, 21 (2), 217-256. Disponible en: http://www.riie.espe.edu.ec.
Aguiar R., Andrade V., Guaygua B., Piedra P. (2016), Aplicación del Método Espectral en los Bloques 6 y 7 con aisladores sísmicos de la UFA-
ESPE en Ecuador, Revista Ciencia, 18 (1), 67-92. Disponible en: http://www.revista_ciencia.espe.edu.ec
ASCE Standard (ASCE/SEI 41), Seismic Evaluation and Retrofit of Existing Buildings, American Society of Engineers, 2013.
ASCE Standard (ASCE/SEI 7-10), Minimum Design Loads for Buildings and Other Structures, American Society of Civil Engineers, Reston,
Virginia.
ATC-40 (1996), Seismic evaluation and retrofit of concrete buildings, Redwood City, California: Applied.
Bonifaz H. (2015), Informe de suelos. Edificio de Investigaciones. Laboratorio de Ensayo de Materiales LEM-ESPE, 33 p., Sangolquí.
Constantinou M., Aguiar R., Morales E., Caiza P. (2016), “Desempeño de aislador FPT8833/12-12/8-5 en el análisis sísmico del Centro de
Investigaciones y de Post Grado”, Revista Internacional de Ingeniería de Estructuras, 21 (1), 1-25. Disponible en:
http://www.riie.espe.edu.ec

Revista Ingeniería de Construcción Vol 32 Nº3 Diciembre de 2017 www.ricuc.cl 129

Chopra A. K. (2001), Dynamic of structures: Theory and aplications to earthquake engineering, 2nd edn. Prentice Hall: Saddle River New York.
Chopra A. K. (2014), Dynamic of Structures Theory and application to earthquake engineering. 4th edition.
Fadi F. and Constantinou M. (2010), Evaluation of simplified methods of analysis for structures with triple friction pendulum isolators,
Earthquake Engineering and Structural Dynamics, 39, 5-22.
Fenz D., Constantinou M., (2007), Mechanical behavior of multi-spherical sliding bearings, Technical Report MCEER-08-0007,
Multidisciplinary Center for Earthquake Engineering Research. Multidisciplinary Center for Earthquake Engineering Research. Available at
http://mceer.buffalo.edu/publications/catalog/reports/Mechanical-Behavior-of-Multi-Spherical-Sliding-Bearings-MCEER-08-0007.html
Fenz D.M. and Constantinou M.C. (2008), Development, implementation and verification of Dynamic Analysis Models for Multi-Spherical
sliding bearings. Report No. MCEER-08-0018, Multidisciplinary Center for Earthquake Engineering Research, Buffalo, NY. Available at
http://mceer.buffalo.edu/publications/catalog/reports/Development-Implementation-and-Verification-of-Dynamic-Analysis-Models-for-Multi-
Spherical-Sliding-Bearings-MCEER-08-0018.html
Ger J. and Cheng F. Y. (2012), Seismic Design Aids for Nonlinear Pushover Analysis of Reinforced Concrete and Steel Bridges. Boca Raton:
Taylor & Francis Group.
Lamar S. and Fortoul C. (2007), Análisis estático de estructuras. Formulación matricial. Editorial CITECI, Caracas, Venezuela.
McVitty W.J. y Constantinou M.C. (2015), Property Modification Factors for Seismic Isolators: Design Guidance for Buildings, MCEER Report.
15-0005, 242 p. Multidisciplinary Center for Earthquake Engineering Research, State University of New York at Buffalo, Buffalo, NY, 2015.
Available at http://mceer.buffalo.edu/publications/catalog/reports/Property-Modification-Factors-for-Seismic-Isolators-Design-Guidance-for-
Buildings-MCEER-15-0005.html
NEC-15 (2015), Norma Ecuatoriana de la Construcción NEC-SE-DS, Ministerio de la Vivienda MIDUVI.

Revista Ingeniería de Construcción Vol 32 Nº3 Diciembre de 2017 www.ricuc.cl

130